INDUSTRIAL AND ENGINEERING CHEMISTRY
154
.
freshly prepared alkali blue ink may be placed upon the glass surface and viewed as illustrated in Figure 8 ; ihe spectrophotometric measurement of the observed yellow color is also shown. On the other hand, if the vie.iying conditions are changed to t'hose of Figure 9, the observed color \Ti11 be green. The spectrophotometric curve of this color is also given. The reason the color changes from yellow to green in the two cases is that the angle a t vhich the light strikes the sample is changed. This experiment may not be of practical importance t o the paint' or ink man, but the various colors exhibited by alkali blue, which include blue, yellow, violet, and green, are all to be expected from a consideratioii of the Fresnel equation. This is evidence of the soundness of the optical explanation of the phemonenoii of interface bronze. CONTROL OF BRONZE
The appearance of bronze can be controlled to some extent by lormulation. For instance, if alkali blue is made up Kith one part of varnish and compared with a second sample made up with thrre parts of varnish, the pulldomil wit'h lesser varnish mill nppear more bronzy. However, since the purpose of this paper is t o tlrfine bronze, give the optical explanations, and present >onie illustrative nieasurements, consideration of the factors of formulation, manufacture, or application by which bronze may he altered is outpide the scope of this discussion.
Vol. 39, No. 2
ACKNOWLEDGMENT
The authors are indebted to C. F. Pinzka for arithmetic computations on interface and int,erfererrcebronze, to v. L, ~i~~ fo,. instigation and interest in t,his invest,igation, to G , 1,. R~~~~ c. nIaresh for the photographs, an(it o c. C. Can&, J . 11.Feild, alld 1-1.l\I&ormacic for preparing some t,he studied, LITERATURE CITED
Cornell ~;'ngineel.,11, 23 (3945), (1) ( 2 Fosa,C, F,, ,I, Optical so,,, Am,, 28, 389 (1938). (3 Godlove, I. H., Inter-Society Color Council, June 27, 1940. (4 Hardy, A. C., Handbook of Colorirrletry, Cambridge, Mass., Technology Press, 1938, ( 5 ) Housmn, R%.. "Treatise on I,ight", London, Longmans, Green. & Co.. 1938. (6) International Printing I n k Corp., Research Lab., "Color as Light" (1933). ( 7 ) AIason, c. J . P h y 8 . Chem., 27, 201, 401 (1923); 28, 1233 (1924): 30, 382 (192G:: 31, 321, 1856 (1927). (8; Port,nian. A, B., A m . I n k X a l z e r , 9, 35 (RIay 1931). (9) ~ oI%.,,I. ~optical ~sot. L4m,, ~ 29, 235 ~ ~ , ;lo) TTilliams, G. C., and Rliiller. N. W., Pn .lIgi'., 19, 240 (1939).
w.,
~ ' R I m E x T R D before the Division of Paint, 1-arnish, and Plastics (IIteiiiistry at t h e 110th hloeting of t h e A V I C R I C . ~ X CHEMICAL SOCIKTY at, Chicago, 111
d
J
A. .k O'KELLY, J. ICELLETT, 4ND J. PLUCKER Socony-Vacuum Laboratories, iPadsboro, N . J .
T h e monaallcylation of benzene with olefins of low molecular weight, over silica-aluniiiia catalysts of the type used in commercial catalytic cracking and at elevated temperatures, is accomplished with substantial yields of monortlliylbeiizeiies at relatively low pressures. The utilization of ethylene was found to be favored by increased reaction time, increased temperature, and increased molar ratio of benzene to ethylene. Small amounts ofpolyeth?-lheiizeues produced maq be recycled in the charge stream to give increased yields of monoethylbenzene based on ethj lene and benzene consumed. The catalyst indicates long Me and sustained activity under the conditions used. i cyclic operation such as is used in catalqtic cracking i s adaptable to the production of monoethylbenzene in which air rcgeneration of the catalyst is carried ont at temperatures in the same range as the reaclion temperatures.
HE low temperature alkylation of aromatic hydrocarbons wit,h olefins in the presence of various catalysts has been liberally described. Ipat,ieff, Corson, and Pines ( 5 )showed that sulfuric acid cat'alyzes the reaction between benzene and amylene t o give good yields of mono-, di-, and tria,mylbenzenes. Propylene and butenes arc also readily reacted with benzene in the presence of sulfuric acid (5, 17). Substantially t h e mine results were obtained by the use of hydrogen fluoride ( 1 6 ) ~Ethylene, however, did not give appreciable yields of et,hylbenzenein either case. The standard method for effecting the reaction of ethylene $th benzene ar. relatively lo^ temperatures to give ethylbensenes mas
dcacribed by Balsohn (1) and ivas the subjuct of numerous invcstigations (2, 3, 4 , 9 , 1 1 ) . The catalyst for the reaction consists mainly of inetal halides of the Friedel-Crafts type. Ipatiefl, Pines, and Kornaremky ( 7 ) used o-phosplioi,ic acid at 300" C. to effect the reaction. Pardee and Dodge (12) dcscritird the use of sodium-aluminum chloride comp1exi.s supported oii pumice, and extruded phosphoric acid-kieselguhr cat,alyst,s,in thci formation of ethylbenzene from ethylem and !)enzenc in t h t : vapor phase at 230-270" C., JTith pressures in tlic older of 200 pounds per square inch. More recently extensive investigatioiis of the use of an extruded phosphoric acid-kieselguhr catalyst in pellet form containing 62-63yc by weight of P20sfor the reaction were reported by Mattox (I@) and Ipatieff (8).The tempoet ur(1 used x a s 270-300" C. ; thisindicat,cd vapor phase operation. Thween benzene and monoethylcatalyst is designated as aged in runs 10 and I1 (Table 11). AIbenzene in the liquid product, distillation (100-132" C.) suggested though the percentage of unreacted ethylene a t the same temperature and spacc rate increased (comparison of runs 2 and VAPOR PHASEPROIJYLATION OF BENZENE (SINGLE-PASS) TABLE111. CONTINUOUS 10) with the aged catalyst, substantially TWIN SYKTHETIC SILICA-ALUMIKA CATALYST the same yields aere obtained by reducRun No. 1 2 3a 45 5a ing the space rate (runs 2 and 11). This Charge, wt. % 88.3 91.5 81.0 81.5 81 .O Benzene is a conservative. indication of catalyst Propylene 11.7 8.5 7.6 11.1 7.4 , 171.. 64 ... ... 11.4 Propane life, since coke deposition is much greater 6.0 Mole ratio, beniene to propylene 4.05 5.78 5.9 5.9 under cracking conditions; therefore the Space velocity liquid, cc. a t 2 2 3 3 3 25' C./cc. catalyst/hr. regeneration conditions are more severe. 422 462 377 Temp. C. 382 382 500 500 76 75 500 Pressure, lb./sq. in. gage The residue of the reachion product in the On-stream time, min. 20 20 30 30 30 Catalyst treatment Fresh Regenerated Fresh Regenerated Regenerated single pass operation contained mainly catalyst from run 1 catalyst from run 3 from run 4 diethylbenzene with small amounts of Products, wt. % of charge (no loss basis) higher et,hylbenzenes, as distillation 75.0 76.2 71.1 81.8 76.8 Benzene (80-100" C.) analysis of combined residues indicated. Intermediate (100-145' C . ) 0.9 0.9 1.0 1.3 2.1 Isopropylbenzene (145This residue would not be lost to mono1 8 . 0 1 3 . 2 11.9 7 . 9 1 4.5 155' C . ) 3.5 3.2 3.4 2.5 3.1 Residue (above 155' C.) ethylbenzene production, since it, is posXi1 8 . 2 Q . 6 1 0 .4 0.7 Unreacted gas 0.6 0.5 0.5 Coke sible t o recycle these polyethylbenxenes 0.8 0.5 Total 1oo.o loo.0 100.0 100.0 with a molar excess of benzene and obtain increased yields of monoethylbenYield monopropylbenzene, % of 56.6 57.8 47.5 68.8 theoretical based on propylene 55.0 zene based on ethylene consumed. This Runs 3, 4, and 5 were conducted with apparatus similar to that described, adapted for use of higher pressures. procedure of alkyl group transfer has been ,eported (4, 16). Run 7 was made to
1oo.o
Q
-
~
1sa
.
INDUSTRIAL AND ENGINEERING CHEMISTRY
However. in reported results (4, 16) iiiicler similar conditions of ('atnlyst and temperature, there was 110 iiiciication of toluene forni:rtiun irom ethyl group decomposition i n thrl ethyl group ~ratisfcrhorweeii polyetliylbenzerieund benzene. I n the product tlistiibut,ioii calculatioii tliis sniall iii terrriediaie fraction was considcrctl to br half benzene and half monoethylbenzene. Sevcral continuous rxpcriiiicmlal rutis w r r ( ~matie with propyl('lie, propylene-propaiic. mixtures, and benzene for bhc formation of cuinciic. Data lor tlicsc runs nppcar i n Table 111. Tcmperature conditiolis for tliis rcacl~iouarc lcss scvcrc than Ihow for the ethylation reaction, and the reaction proceeds with greater ease. Runs 3 , 4 , arid 5 indicate that subst'antialselecbive reaction occurred in the presence of paraffin gases which were inert under lhe condit,ions of the experiment,.
Vel. 39, No. 2
(3) Grosse and Ipatieff, J . Org. Chem., 1, 559 (1936). (4) Hansford, Myers, ani1 Sac'lianell, I N D . IGNG. I,, Bull. SOC. chim., [2] 31, 539 (1879). (2) Franris and Reid, TND. EN^. CHSM.,38, 1194 (1946).
PRESENTED before the Division of Petroleum Chemistry at the 109th Rleet. \ \ I T ~ R T C A S C I I ~ ~ I IFi)crr.~ru, C ~ ~ T .Ailnrrtir, City, IC,J.
ing 01 t h c
0
d
KEYSOLD
c. hIERPBILE
Philadelphia Quartz Company, Philadelphia, Pa.
-A
phase study of commercial mixed soap-water-electrolyte systems w a s made using sodium rhloride and nine salts industrially important as soap builders. The data cover soap concentrations to 50%. electrol! te conceutraThe salts used tions to 27%, and temperatures to 180" C:. were aodium chloride, carbotiatr. and Letraborate, trisod i u m phocphate, tetramdiurri p~rophosphatc, Calpon (sodium heuamelaphosphate), sodium metasilicate, sodium silicates of Si02,"azO ratios by wright o f 2.46 and
HE valur of ztddirig alkaline cileclrolytes or builders, such as the soluble silicates, to soaps is attest,ed by many years of experience and numerous publications ( I , S,5, 6,%5,86).In 1037, the last year for which accurate data on consumption are npparently availahlc, 160,000 tons of 10"lhumi? sodium silicates and 39,500 toils of sodium phosphates merc consumed hy thc United States Poap industry according t,o t'he Buresii of the Census ("8). Since then the amoimt,s of such builders used have substantially iircrcased. \Vart,iine shortages or fats, oils, and rosins for soapmaking re-cinphasized t,lie value of builders in soap. Attention has heeii focused mi the amouiits of various builder6 which call lie added to different s o a p iiiidcr w r y i n g conditions and on the nat,ure of such sy During the past, twenty-five years t,he m:iiiuEact,lire o f soap, previously a n industrial art,, has been studicd scientifically from the phase-rule point of vii: by McRain and collaborators (13, 16, 18),Ferguson airti lticl grams of approximnt~clyit do wit,h sodium or potiissium rhloride ha,vo bten piiblislied (7, I S , 1 8 ) . Only incomplct,e data on t,tit. behavior of soaps a i t > l iother salts are available. Probably t h o earliest, work was: (lone iii 1888 by Hofmeister, who studied the salting out of sodium oleatt from solution by several salts ( f U ) . Later workers ( d , I d , $1, $2,
3.93, aiid a potassium silicate of SiOziI620 ratio h y ~ t r i g h l of 2.01. The solubility of the soap i n solutions of these salts and their effect on the transition from crystalline to liquid cry s t d i n e soap varies widely both on weight and iiiolecular bases. The order of increasing d e c t difrers H ith coiicenLration and temperature; howevei, suflicient regularities exist t o enable predictions to be made of the phase diagrams for other soaps and at other concentrations.
Z/i) measured the amounts of sodium and potassium chloride, carbonate, and hydroxide required to salt out various single and mixed soaps. These data vere summarized and interpreted by McBain and Walls (20). Qualitative observations on gelation of soaps in the presence of sodium carbonate, borate, and silicate were reported by Fischer ( 9 ) . McBain and Pitter (17) measured the concentrations of eleven electrolytes required to salt out 0.25 weight normal (approximately 6.5%) sodium palmitate. McBain, Vold, and Gardiner (18) found that substit,uting for sodium chloride a sodium silicate, having a Si02/?5a20 weight ratio of 3.18 in concentrat,ions of 3.6 amid 5.6c4, caused lo?', sodium oleate 10 set to an elastic, transparent jelly instead of a liquid crystal. I f a determination is made of the relative efficiency of various d t , s in salting out soap a t a single concentration and tcmperature, hypothetical phase diagrams can be outlined for aqueous tems of that soap with all of these salts. The diagram for one of them must he known; the assumption is then made that the relative efficiency is the samc for different soap conce involving several crystallint: and liquid-crystalline 1 different temperatures. Tliis paper reports a study of the phase beliavior of a typical cxnnmercial mixed soap with one potassium and nine sodium salts. The salts used were sodium chloride, carbonate, and
